Sanjay Kaul, MD: Discusses the NY Times article on Crestor risks | CurrentMedicine.tv

March 31, 2010

The NY Times featured a front-page article on March 31 discussing the risks of expanding the indication for the statin Crestor to people with normal levels of cholesterol. This story came on the heels of another story appearing in Time magazine that questioned statin risks and benefits in women. We asked cardiologist Sanjay Kaul for his thoughts.

Dr. Kaul: Although statins have been shown to increase the risk of incident diabetes, the key question to ask is whether their favorable impact on cardiovascular outcomes are reduced in those that are rendered diabetic by statin therapy. It is a paradox that interventions that increase incident diabetes (such as thiazide diuretics, beta blocker, statins, even niacin) improve cardiovascular outcomes. In contrast, none of the interventions (with the possible exception of metformin) that reduce incident diabetes or improve glycemic control (such as TZDs, nateglinide, insulin, sulfonylureas,

GLP-1 analogs, etc.) have been shown to reduce cardiovascular outcomes.

I think the broader message of the NY Times article is on target.

Lifestyle modification trumps pharmacologic intervention for primary prevention.

With regards to the Times magazine story, a treatment benefit of statin, be it total cardiovascular disease or total mortality events, that is not statistically discernible from null in nearly 20,000 subjects, does not provide robust evidence of benefit in favor of statins in women. The total cardiovascular disease benefit observed with statins in the exclusively primary prevention trials is driven by the results of JUPITER, which is likely to be an overestimate given the premature stoppage of the trial. Furthermore, the 5-year numbers needed to treat (NNT) estimates in JUPITER are unlikely to be reliable. The trial was stopped at 1.8 years. So, how can one simply extrapolate the NNT to 5 years and assume that the benefits observed at 1.8 years are going to persist at 5 years? People who insist on extrapolating NNT to time frames beyond the actual period of follow-up should be reminded of the perils of random variation, especially when the trials are stopped prematurely.

Many experts have disputed the facts reported in the Times magazine story and have expressed concerns that the story will end up discouraging women to control their risk factors. I did not see anything in that report that is not factual or that encourages women to quit controlling their risk factors. Instead of criticizing the article, we should acknowledge the holes in the evidence that the Times report rightly identifies and generate data to shore up the evidence.

via currentmedicine.tv

Posted via web from Jack's posterous